Hi, I have a computational proteomics application that builds phylogenetic
trees and multiple-sequence alignments. However, the algorithm is extremeley
slow for large sets and I plan to parallelize it using MPI. I am thinking
this will involve a portion that is embarassingly parallel (such as doing
the all-against-all scoring) and a portion that is somewhat more clever (
like branch-and-bound tree construction)
I have already implemented a working sequential version in C++. Please let
me know if you're interested!
Thanks,
Michael Tung